This retrospective comparative study was approved by Medical Research Ethics Committee of our hospital. From Jan 2019 to May 2020, a total of 91 cases were diagnosed as clavicle fractures in our institution. Patients with midshaft clavicle fractures were included and those corresponding to the exclusion criterion were excluded: 1) a patient’s refusal to participate; 2) proximal or distal clavicle fracture; 3) neurovascular injury; 4) open fractures; 5) pathological fractures; 6) fracture seen beyond 2 weeks after the injury; 8).intramedullary nailing fixation. Finally, 58 cases were retrospectively reviewed in this study. According to the surgical technique they were divided into minimally invasive group and open reduction group(Fig.1).
Minimally invasive plate osteosynthesis
Under general anaesthesia, the patient was placed in a supine position on a radiolucent operating table. For ease of closed reduction muscle relaxant was always required in our cases. A well-padded adequately sized soft bump was placed between the scapulae. To make sure there was no shelter when the C-arm was used, and was tilted in order to obtain inlet and outlet views, if necessary. The involved shoulder and the whole upper extremity were prepared and draped in a sterile fashion so that the whole upper extremity was placed in the operative field. This allowed easier intra-operative manipulation and reduction.
After palpating, the proximal end, distal end and fracture site, a suitable distal clavicle anatomic locking plate was selected and placed along the clavicle. The provisional position of the plate was drawn on the skin under an image intensifier（Fig. 2.c）. According to the provisional position of the plate, a 1.0cm-sized oblique incision was performed along the long axial of the clavicle at the center of the distal clavicle end, sharp dissection to the bone and exposing the distal clavicle end. A sub-muscular tunnel along the clavicle was developed with a periosteal elevator and the plate was inserted from the distal incision.
A locking sleeve was placed at the distal end of the plate. The distal plate and segment were fixed in the optimal position with both anterior borders of the plate and distal clavicle end at same horizontal using a K-wire through the locking hole. A tractive force was used through the locking sleeve. After restoration of clavicular length and realignment of the comminuted fragment, we percutaneously fixed the proximal plate and fragment using K-wire through a proximal locking hole to maintain the realignment (Fig. 2.d), and the status of the reduction was confirmed by fluoroscopy (Fig. 2.e).
In some cases the fracture ends were still severely displaced in a “back to back” position after preliminary reduction(Fig. 3.bd). Before inserting the plate we could insert a small hook from the distal incision along the sub-muscular tunnel and fix the hook in the medullary canal of the distal fragment, then using a traction or rotate force to reduce the displacement (Fig. 3.ef), this technique was reported in our previous reach.If the gap between distal clavicle and plate was wide after inserting the plate, we could use a conventional screw first to reduce the gap when we fixed the distal plate and segment (Fig. 3.gh). A conventional screw also could be used to reduce the proximal fragment displacement or minimize the gap between the plate and proximal fragment.
In some cases the sharp vertical segment could insert into the surrounding soft tissue. If there was still severely displaced after preliminary reduction(Fig. 5.b). We could insert forceps from the distal incision along the sub-muscular tunnel to clip the vertical segment, then using a traction and rotate force to unlock it (Fig. 5.cde).
The same size and type distal locking plate was used as a guide in proximal percutaneous fixation once an acceptable realignment of the fracture was achieved. A 0.5cm-sized incision was made on proximal site and fixed the proximal segment with the plate using a conventional screw (Fig. 2.f). The proximal K-wire was retrieved with the slide of the skin on clavicle region the proximal segment was fixed using a percutaneous technique through two 0.5cm-sized incisions, and at least three 3.5mm screws were fixed at proximal side of the fracture. The distal segment was fixed using 4-6 2.7mm locking screws.
Following the placement of distal and proximal screws, intra-operative ﬂuoroscopic images were taken again to confirm the good reduction and fixation. Upon satisfactory radiograph, the wound was irrigated, the proximal incisions could be sutured using intradermic suture technique and the lateral incision was sutured in the standard fashion without a drain (Fig. 4.b, Fig. 5.i). Fluoroscopy times and operation time were also recorded.
Open reduction and plate fixation
Under general anaesthesia, the patient was placed in a supine position on a radiolucent operating table. An incision along the long axial of the clavicle was made after draping and preparation. Supraclavicular nerve was identified and protected. Sharp dissection to the bone and expose the fracture site, the fracture site was reduced and fixed with K-wires after debridement. A good reduction was confirmed by intraoperative ﬂuoroscopy. A suitable distal clavicle anatomic locking plate was selected and fixed the fracture site with both proximal and distal segment at least three screws. Intra-operative ﬂuoroscopic images were taken again to confirm the good reduction and fixation. The incision was sutured using standard fashion without a drain (Fig. 6).
Assessment of outcomes
Baseline characteristics, including age, sex, side, mechanism of injury, Robinson classification, interval from injury to surgery, and follow-up period, were assessed. The fracture pattern was classified according to Robinson classiﬁcation system .
Clinical and radiological outcomes of the technique included reduction quality, union time, and complications. For clinical evaluation, we used the Constant-Murley score , which were assessed from pain, daily activities, range of motion and power at sixth month postoperative. Incision length was measured at sixth month postoperative, and the incision length was proximal incision length plus distal incision length in minimally invasive group. The quality of reduction was measured as the proportional difference in clavicle length between the injured and uninjured side . An anteroposterior view X-ray was taken at the latest follow-up to reveal shortening of the clavicle. The radiographs were examined for evidence of fracture healing, short displacement or implant failure. Fracture healing was defined clinically and radiographically as the absence of pain and visible callus on anteroposterior X-ray plain. Radiographs were read by an independent examiner blinded to the study details in order to verify the short displacement and state of the bone union. Complications, such as nonunion, postoperative infection, and major neurovascular injury, were assessed.
Statistical analyses of the data were performed by an independent statistician blinded to clinical outcomes using the Statistical Package for the Social Sciences (SPSS), version 20.0 (SPSS, Inc., Chicago, IL, USA). All quantitative variables were expressed as means and standard deviation (SD) and paired student’s t tests were used to analyze the difference. Categorical variables were shown as number and percentages (%) and tested by the chi-squared test. Statistical significance was defined at the level of P < 0.05.